Sure, but the lock isn't needed at all here, so while your fix is
probably correct for other cases that actually need a lock call, this
isn't one of them.
-Rasmus
Frank M. Kromann wrote:
> Hi Rasmus,
>
> But the definition of flock in win32/flock.c() is identical to the
> definition of php_flock
Hi Rasmus,
But the definition of flock in win32/flock.c() is identical to the
definition of php_flock() in ext/standard/flock_compat.c.
I don't see any reason to hae the same function defined twice with two
different names.
- Frank
> Frank M. Kromann wrote:
> > Hello Everyone,
> >
> > I hope I
Frank M. Kromann wrote:
> Hello Everyone,
>
> I hope I'm not too far off here but getting rid of flock.c in
> win32\build\config.w32 an adding flock_compat to ext\standard\config.w32
> seams to solve the compile problems, at least in the 5.2 branch.
Rob tracked it down to the fact that it was usi
Hello Everyone,
I hope I'm not too far off here but getting rid of flock.c in
win32\build\config.w32 an adding flock_compat to ext\standard\config.w32
seams to solve the compile problems, at least in the 5.2 branch.
- Frank
--
PHP Internals - PHP Runtime Development Mailing List
To unsubscribe,
No difference using sprintf()/fwrite() instead of fprintf().
I did come across a similar issue from apache:
http://mail-archives.apache.org/mod_mbox/httpd-dev/199503.mbox/[EMAIL PROTECTED]
Changing to use VCWD_OPEN_MODE, write() and close() seems to work.
Rob
Rasmus Lerdorf wrote:
That still
That still doesn't make sense to me. Could you try doing an sprintf()
followed by an fwrite() then? Is fprintf() perhaps doing something
stupid internally? I'd really like to avoid a lock here.
-Rasmus
Rob Richards wrote:
> I can confirm that it does fail with the single fprintf call without t
I can confirm that it does fail with the single fprintf call without the
lock on Win32.
Using flock() does fix the problem in my tests.
I was able to fix win build using flock() instead of php_flock() and
defining HAVE_FLOCK in win32/flock.h
Rob
Richard Quadling wrote:
This test is with the
Jingcheng Zhang wrote:
Hi internals,
I used to use abstract static function prior to php 5.2, because when I
consider static class as a singleton, I can use abstract static function to
delay the implementation of some methods to the derived classes. This
happens, for example, when I declare a
This test is with the multiline fprintf and no lock.
Adding the php_flock() has stalled the build process for Win32 (I
can't build on Win32 as I don't know how!).
On 05/04/07, Rasmus Lerdorf <[EMAIL PROTECTED]> wrote:
Yes, but again, is this test with the single fprintf call? That's the
real
Is it possible that at the lowest layer within the C library that 4K
blocking is performed?
On 05/04/07, Rasmus Lerdorf <[EMAIL PROTECTED]> wrote:
Ilia Alshanetsky wrote:
> Rasmus,
>
> Sorry for the delay in the reply. According to my tests on linux using
> the sample script provided by the orig
Yes, but again, is this test with the single fprintf call? That's the
real fix for this problem, not the lock.
-Rasmus
Richard Quadling wrote:
> Using PHP 5.2.2-dev (cli) (built: Mar 23 2007 07:02:57) I can
> replicate the problem on Windows.
>
> Using this single line command at the CMD prompt
Ilia Alshanetsky wrote:
> Rasmus,
>
> Sorry for the delay in the reply. According to my tests on linux using
> the sample script provided by the original bug reporter having no lock
> causes a problem when the error message is >4k in length. In this case
> multiple buffers are used and corruption
Using ProcMon on WinXP, you can see that the different processes
interleave the writes but in blocks of 4096 bytes. The remaining 929
bytes will sometimes be written after another processes 4096/929
bytes.
Locking IS required here.
On 05/04/07, Richard Quadling <[EMAIL PROTECTED]> wrote:
Using
Using PHP 5.2.2-dev (cli) (built: Mar 23 2007 07:02:57) I can
replicate the problem on Windows.
Using this single line command at the CMD prompt:
for %x in (A B C D E F G H I J K L M N O P Q R S T U V W X Y Z) do
start php -r
"ini_set('error_log','/tmp/test.log');for($i=0;$i<100;$i++)error_log(
On 4/5/07, Tijnema ! <[EMAIL PROTECTED]> wrote:
Hi,
Is the mail server of the PHP general list down?
It keeps bouncing my mesages with some SMTP error.
Tijnema
To clarify myself, the error i get is:
PERM_FAILURE: SMTP Error (state 12): 550 we're manly enough already
It seems that it only ha
Hi,
Is the mail server of the PHP general list down?
It keeps bouncing my mesages with some SMTP error.
Tijnema
--
PHP Internals - PHP Runtime Development Mailing List
To unsubscribe, visit: http://www.php.net/unsub.php
Rasmus,
Sorry for the delay in the reply. According to my tests on linux
using the sample script provided by the original bug reporter having
no lock causes a problem when the error message is >4k in length. In
this case multiple buffers are used and corruption can happen (it did
on a dua
http://rquadling.php1h.com/main.c.diff.txt for HEAD. I don't know how
to patch a specific version.
On 05/04/07, Rasmus Lerdorf <[EMAIL PROTECTED]> wrote:
Matt Wilmas wrote:
> Hi,
>
> Maybe just a Windows problem if it wasn't noticed yet, but I was compiling
> the latest 5.2 snapshot and got:
>
18 matches
Mail list logo
